Numerous methods have been developed for inferring (reverse engineering) gene regulatory networks from expression data. However, both their absolute and comparative performance remain poorly understood. The aim of this project is to provide benchmarks and tools for rigorous testing of methods for gene network inference.
We have developed novel approaches for the generation of realistic in silico benchmarks and for the identification of systematic errors of network inference algorithms. Our framework is available as an easy-to-use Java tool called GeneNetWeaver (GNW). We are using GNW to provide the DREAM in silico challenge, an annual community-wide network inference challenge that is currently the most widely used benchmark suite in the field.
[13 Jan 2010]
The new version 2.0 of GeneNetWeaver is released.
[16 Jun 2009]
The LIS releases the gene-network inference challenges of the DREAM4 conference, to be held in December 2009 at the Broad Institute of MIT and Harvard.
[21 Oct 2008]
About 30 teams participate in the In Silico Challenges provided by the LIS for the 3rd DREAM Conference at MIT.
[26 Nov 2007]
DREAM2 gene network reverse engineering challenges: the biomimetic approach based on Analog Genetic Encoding (AGE) is best performer in the five-gene network challenge.